A Modified Image Reconstruction Algorithm Based on Compressed Sensing

Abstract

Compressed sensing theory is a new kind of making full use of signal sparsity or compressible sampling theory. The theory suggests that collecting a small amount of signal values can realize accurate reconstruction of sparse or compressed signal. Through the research and summary of the existing reconstruction algorithm, the paper proposes a new adaptive matching pursuit algorithm based on regularization Regularized Adaptive Matching Pursuit (RAMP) for compressed sensing signal reconstruction, called blocking sparsity adaptive regularized matching pursuit (BSARMP) algorithms. In order to reduce the scale of a single observation matrix processing and the single processing speed, a novel method based on image blocking is presented in this paper, thereby improving the overall running time.